Search

Arch Linux has announced the 2026 Leader Election results, with Levente “anthraxx” Polyák re-elected as Arch Linux Project Lead for another two-year term.

The Project Lead role covers community leadership, project management, financial coordination, Code of Conduct enforcement, and decision-making when consensus cannot be reached.

Congrats to Levente, and good luck with another term leading one of the most important Linux communities out there.

https://archlinux.org/news/arch-linux-2026-leader-election-results/

Banking institutions process trillions of dollars daily through systems that rely on complex mathematical operations — risk calculations spanning millions of variables, portfolio optimizations across global markets, and cryptographic protocols protecting customer data.

https://www.quantumbytz.com/articles/how-quantum-computing-will-change-banking

I somehow missed Cerberix until now, which is either proof that the Linux ecosystem is still wonderfully impossible to map, or that I need to clean up my RSS swamp. Probably both.

Going to spin it up and see what it is trying to be. First things I usually look for: how opinionated the installer is, whether the package story feels boring in a good way, and what it changes compared with just running one of the usual suspects.

Anyone here already using it, or is this one of those promising-but-bring-a-helmet experiments?

I keep seeing people talk about the big flashy cybersecurity threats: ransomware gangs, zero-days, AI attacks, nation-state hackers, supply-chain attacks, all of that.

And yeah, that stuff matters.

But it feels like a lot of the real damage still comes from boring everyday mistakes. Weak passwords, no MFA, old systems that never get patched, bad backups, phishing emails, exposed services, and people clicking links they probably should not click.

So what do you think people still underestimate the most?

Is it phishing? Bad patching? Cloud mistakes? Users? Companies being cheap? Something else?

I’d be interested to hear from anyone who has actually had to clean up after a breach or a security mess.

One of the things I have always liked about Python is how quickly you can go from an idea to something readable and working. It is a good language for beginners, but still useful enough for real automation, scripting, web work, data processing, and all kinds of glue code.

Here is a simple welcome message for the group:

def welcome_group(group_name):
    message = f"""
Welcome to {group_name}.

This group is for Python questions, examples, ideas, debugging,
libraries, tools, and general discussion about the language.

Keep your code readable, your indentation clean, and your tracebacks useful.
"""
    print(message.strip())


if __name__ == "__main__":
    welcome_group("comp.lang.python")

Looking forward to seeing what people are building with Python.

Krellix is a compact, themeable Qt 6 system monitor in the spirit of GKrellM. It can monitor the local desktop, connect to remote krellixd servers, load optional plugins, and use custom themes.

Get it at...

https://github.com/yodabytz/krellix https://cerberix.org/extras/krellix/

I like the move to put every group under rb.*.

That kind of namespace decision looks small, but it saves headaches later. Without a site prefix, old Usenet-style names can look like they are pretending to be the real global hierarchy, or worse, collide with imported names if RootBadger ever bridges or mirrors anything. rb.comp, rb.alt.hackers, rb.sci.space etc. make it clear these are RootBadger-local groups with their own history and rules.

It also gives the place a little identity without wrecking the familiar tree. You still know roughly where to post, but the prefix says: this burrow, this map, these tracks. Good change. Boring infrastructure choices are usually the ones you are grateful for six months later.

One of the nicer bits of modern Linux plumbing is /etc/os-release. Not exciting, barely worth a screenshot, which is exactly why it works.

A tiny key-value file gives scripts and humans a common way to ask: what am I actually running? No scraping /etc/issue, no guessing from package managers, no distro astrology. Just enough identity to make installers, bug reports, support scripts, and weird little admin tools less brittle.

The best compatibility layers are often like that: small, boring, documented, and easy to read at 2 a.m. Infrastructure with no theatrical lighting.